Applied Scientist - Embodied AI for Autonomous Driving

Wayve; Sacramento (CA); Hybrid; USD 312,000 - 370,000; Full time

Job summary

Wayve is the leading developer of Embodied AI technology focused on autonomous driving and foundation models. This full-time role is based in Sunnyvale with a hybrid policy, and involves building and evaluating large-scale pretrained models, data curation, and advancing the GAIA/LINGO ecosystems to drive safer, smarter driving systems.

The role emphasizes data wrangling at scale, training foundation models, and rigorous offline/closed-loop evaluation, collaborating across teams to push

Qualifications

• Masters with 6+ years of relevant experience, or a PhD with 2+ years, in computer science, ML, robotics, mathematics, or related field.
• Strong ML and software fundamentals with a track record of production ML at scale.
• Hands-on strength in data curation, foundation-model training, large-scale data wrangling, and evaluation.
• Experience with large-scale data and large neural networks.
• Fluency in Python and a modern deep-learning framework (PyTorch or similar).

Responsibilities

• Mine world-scale fleet data for rare, safety-critical events using active learning and embedding-based retrieval.
• Figure out good training data mix and curate repeatable processes across cities and sensor rigs.
• Build high-quality enrichments and labeling pipelines at scale.
• Build and fine-tune large-scale pretrained models and run experiments to de-risk ideas.
• Help build best embodied VLM/VLA for driving (LINGO line).
• Design rigorous offline and closed-loop evaluation with coverage of rare scenarios.
• Use GAIA for counterfactual evaluation safely and at scale.
• Contribute across the foundation-model stack (GAIA, policy learning, RL, reward modeling).

Skills

Python

PyTorch

Data wrangling

Foundation models

Evaluation

Large datasets

Education

Master's degree or PhD

Tools

Milvus

Ray Data

Daft

Spark

Job description

Wayve is the leading developer of Embodied AI technology focused on autonomous driving and foundation models. This full-time role is based in Sunnyvale with a hybrid policy, and involves building and evaluating large-scale pretrained models, data curation, and advancing the GAIA/LINGO ecosystems to drive safer, smarter driving systems.

The role emphasizes data wrangling at scale, training foundation models, and rigorous offline/closed-loop evaluation, collaborating across teams to push